Lieke Asma

There Is No Free Won’t

The Role Definitions Play
For personal use only -- not for reproduction

Abstract: In this paper, I analyse how neuroscientists come to the con-

Copyright (c) Imprint Academic 2017

clusion that the brain ‘decides’ what we will do. I do so by focusing

on a recent study on free won’t, from which it is concluded that the
decision to veto is not free. First, I argue that assumptions about
voluntariness and freedom that underlie this and other Libet-style
experiments are more stringent than assumed by other critics (Mele,
2009; Schlosser, 2014). Second, I claim that these assumptions lead to
an experimental setting in which the conclusion that the brain
‘decides’ is almost unavoidable. This is because the only decisions
subjects are allowed to make in these experiments are spontaneous
decisions that are not based on reasons. Even if the subjects acted for
reasons or deliberated about what to do it would not be interpreted as
such. Because of this, alternative definitions of voluntariness and
freedom are necessary for neuroscientific experiments to provide a
valuable contribution to the debate on free will.
Keywords: neuroscience of free will; veto decision; free won’t; Libet-
style experiments; conscious will.

Introduction
Libet-style experiments have received much attention in both neuro-
science and philosophy. In general, these experiments are taken to
show that if subjects are asked to freely decide when to move or what
movement to make, the conscious intention or urge to move and the
actual movement can be predicted on the basis of preceding uncon-
scious brain activity (e.g. Haggard and Eimer, 1999; Libet, 1985;

Correspondence:
Lieke Asma, VU Amsterdam, The Netherlands. Email: l.j.f.asma@vu.nl

Journal of Consciousness Studies, 24, No. 5–6, 2017, pp. 8–23

 THERE IS NO FREE WON’T 9

Soon et al., 2008). Based on these findings, researchers have con-

cluded that the brain prepares or ‘decides’ to initiate the movement
before we are aware this process is taking place, and that the ‘con-
scious self’ is not the ultimate source of these actions (e.g. Caruso,
2012, p. 196). Although the findings suggest that we do not con-
sciously initiate our actions, we might still be able to consciously veto
this initial decision and make a free decision about whether to move.
That is, we may still have free won’t. In recent years, several experi-
ments have been conducted to study this veto decision (Filevich, Kühn
and Haggard, 2012; 2013; Kühn, Haggard and Brass, 2009; Parkinson
and Haggard, 2014). One of these studies was concerned with the
question of whether this veto decision is free, and the results suggest
For personal use only -- not for reproduction

that it is not (Filevich, Kühn and Haggard, 2013). Even if we decide to

refrain from acting, it seems that the brain has made this decision, and
Copyright (c) Imprint Academic 2017

not ‘I’.
To assess whether these findings really pose a threat to free will, it
is important to critically examine the definitions used in these experi-
ments. Neuroscientists typically define and operationalize voluntari-
ness first, and on the basis of the experimental results conclude
whether the voluntary action or decision was free as well.1 This is also
the case for Filevich, Kühn and Haggard’s (2013) experiment: sub-
jects have to voluntarily decide to veto or not and, on the basis of the
occurrence of unconscious neural precursors, it is concluded whether
the decision was also free. Both these definitions are identical in all
Libet-style experiments and have been criticized before (e.g. Mele,
2009; Schlosser, 2014). However, by analysing the free won’t experi-
ment I will show that the conditions for voluntariness and freedom are
more stringent than these philosophers take them to be. I will argue
that because the voluntary decision is (implicitly) operationalized as
not caused by external and internal factors and freedom is understood
as conscious initiation, the experiments are set up in such a way that
the conclusion that the brain ‘decides’ is almost unavoidable. This is
because the only decisions subjects are allowed to make in these

1 This distinction is for example made in Haggard (2008). Voluntary action and deciding
(also referred to as ‘intentional’ or ‘self-generated’, or ‘intentional inhibition’ in the
case of the veto decision) are defined in the neuroscientific literature without referring
to the causal role of consciousness, but by pointing out the role of internal and external
causes and the experience of control (e.g. ibid., p. 934). Whether these actions and
decisions are free depends on the role of conscious processes (e.g. ibid., p. 941, or
Haggard and Libet, 2001, p. 47).
 10 L. ASMA

experiments are spontaneous decisions that are not based on reasons.

Even if the subjects acted for reasons or deliberated about what to do
it would not be interpreted as such. Because of this, alternative defi-
nitions of voluntariness and freedom are necessary for neuroscientific
experiments to provide a valuable contribution to the debate on free
will.
The paper is structured as follows. After explaining the free won’t
experiment by Filevich, Kühn and Haggard (2013) in more detail, I
elaborate on the two underlying assumptions in the third and fourth
sections of the paper: the definition of voluntary deciding and the
relationship between conscious processes and freedom. Next, I sub-
stantiate my claim that the conclusion that the brain ‘decides’ is
For personal use only -- not for reproduction

almost unavoidable. In the final part of the paper I focus on alternative

definitions of voluntariness and freedom.
Copyright (c) Imprint Academic 2017

The Libet-Style Experiment on Free Won’t

Neuroscientists in the field take voluntary action to have three compo-
nents: when to act, what action to perform, and whether to act on the
initial intention or urge (Brass and Haggard, 2007). The latter compo-
nent, the voluntary veto decision, is what the study by Filevich, Kühn
and Haggard (2013) is concerned with.2 This veto decision is
associated with self-control and is related to ‘last-minute inhibition of
an action that has been prepared and is ready for execution’ (ibid., p.
1). Previous studies have shown that both the when and what compo-
nent of voluntary action can be predicted on the basis of preceding
unconscious brain activity (e.g. Haggard and Eimer, 1999; Libet,
1985; Soon et al., 2008). This is taken to be problematic for our
common conception of free will, because it seems the ‘conscious self’
is not the ultimate source of these actions (e.g. Caruso, 2012, p. 196).
Filevich, Kühn and Haggard (2013) aimed to find out whether this
voluntary veto decision can be predicted on the basis of unconscious
brain activity as well. If this would be the case the decision to veto is
not free either. Then, no aspect of our voluntary actions would be truly
‘up to us’.

2 In fact in Filevich et al.’s (2013) study subjects did not have to refrain from acting
altogether, but had to decide to either act immediately or to delay their response — for
now I will take this to be a voluntary veto decision and I will not take into account the
question of whether this is a genuine veto decision.
 THERE IS NO FREE WON’T 11

The experiment by Filevich, Kühn and Haggard (2013) consisted of

five conditions: instructed rapid, instructed delayed, free-choice rapid
(voluntary decision not to veto), free-choice delayed (voluntary
decision to veto), and no-go trials. In the instructed rapid condition
subjects had to press as quickly as possible, and in the instructed
delayed condition they had to press with the ‘shortest possible delay’.
More important are the two free-choice conditions, in which the sub-
jects were not instructed whether to act rapidly or delayed; they were
instructed to choose for themselves what to do. In these free-choice
conditions, subjects were asked ‘to produce roughly 50% rapid and
50% delay responses’ (ibid., p. 4). Furthermore, the subjects were
discouraged from taking the decision at the onset of a trial, because in
For personal use only -- not for reproduction

that case it would not be a veto decision. Subjects first had to acquire
the urge or intention to move and actually prepare to make the move-
Copyright (c) Imprint Academic 2017

ment. The later decision whether to act on this urge or intention to

move should be spontaneous and not prepared in any way.3
The goal of the experiments was to find out whether on the basis of
preceding neural activity this decision to veto or not in the free-choice
conditions could be predicted. By comparing ERPs for free delayed
and free rapid responses, Filevich, Kühn and Haggard (2013) found a
significant difference in neural activity preceding the instruction and
the decision to act either rapidly or delayed. That is, when the subjects
voluntarily chose to respond rapidly stronger preparatory activity was
found compared to their decision to delay. From this the researchers
concluded that in the free-choice conditions the conscious ‘I’ did not
initiate the decision either, since their results identified ‘a candidate
unconscious precursor of the decision to inhibit action’ (ibid., p. 10).
Thus, even though conscious awareness of the intention or urge to
move has played some role in the subsequent decision that is made,
Filevich et al. take the voluntary veto decision to be initiated by the
brain and, therefore, not free.4

3 Mele (2008) analysed Brass and Haggard’s (2007) study and concluded that it is not
clear what strategies participants are actually using in the experiments and, because of
that, whether the participants are really deciding not to act on an intention they
acquired. For the purposes of this paper, I assume that the experiment actually targets
veto decisions and that subjects are able to do and are in fact doing what is asked of
them. My goal is to critically examine the assumptions that play a central role in the
experiment and in the conclusion that is drawn.
4 A further problem that should be pointed out here is that the experiment, as well as the
regular Libet-style experiments, only establishes temporal precedence and the findings
do not allow for claims implying causal determination (see for example Fischborn,
 12 L. ASMA

In the next two sections of this paper I focus on the definitions of

voluntariness and freedom that are central in this experiment and the
conclusion that is drawn.

Voluntary Deciding:
Internally Generated or Underdetermined?
In this section, I focus on the definition of voluntariness that is used in
experiments on the veto decision. In his 2014 paper, Schlosser
addresses this definition in which a, according to him problematic,
distinction is made between internal and external factors that influence
the action and decision. Even though this is the way neuroscientists
explicitly define voluntariness, I show in this section that deciding
For personal use only -- not for reproduction

voluntarily is in fact implicitly operationalized by excluding all possi-

Copyright (c) Imprint Academic 2017

ble factors that may play a role from the experimental setting. Because
of this, making a voluntary veto decision amounts to subjects deciding
without being allowed to take any factor or reason into account.
In neuroscientific experiments on the veto decision, Libet’s original
definition of voluntary action is taken as a starting point. As he states:
(a) [the voluntary action] arises endogenously, not in direct response to
an external stimulus or cue; (b) there are no externally imposed
restrictions or compulsions that directly or immediately control sub-
jects’ initiation and performance of the act; and (c) most important,
subjects feel introspectively that they are performing the act on their
own initiative and that they are free to start or not to start the act as they
wish. (Libet, 1985, pp. 529–30)
Also in these more recent studies on the veto, internal generation is
central: ‘the process or signal that cancels or inhibits the action is not
the result of any external signal or instruction, but is generated
internally by the subjects themselves’ (Filevich, Kühn and Haggard,
2012, p. 1108). Furthermore, neuroscientists in the field link the veto
decision to self-control, i.e. the ability to postpone the immediate
gratification of a desire to act to achieve other benefits, which, in the
case of a voluntary decision, should not be the result of a response to
an external cue but generated internally (Kühn, Haggard and Brass,
2009, pp. 2834–5).5

2016). Since in the current paper my aim is to account for the occurrence of this
temporal precedence, I will not take this problem into account.
5 Examples that are used include refraining from sending an angry email (Filevich, Kühn
and Haggard, 2012), overcoming cigarette addiction (Kühn, Haggard and Brass, 2009),
 THERE IS NO FREE WON’T 13

Thus, a decision to veto should not be directly caused by an external

cue, the process that inhibits the action should be generated internally,
and the person should experience control in order for the decision to
be voluntary. In the remainder of the paper I will refer to this explicit
definition of the voluntary veto decision as internally generated. Of
course it is still quite vague what ‘directly’ means here, or ‘generated
internally’. Unfortunately, this is not specified in more detail by the
researchers. From their discussions it is most clear what it is not; the
decision not to act on an initial intention should not be the direct result
of an external cue (e.g. a red cross), but should be something I decided
for myself. This definition still allows for many different ways in
which a voluntary decision to veto can be made: it could be a decision
For personal use only -- not for reproduction

based on my beliefs, desires, goals, or perhaps because it is my

spontaneous decision to do so. For example, I can decide to refrain
Copyright (c) Imprint Academic 2017

from reaching for a cup of coffee, because (a) I feel I do not want
more coffee, (b) I decided to drink less coffee a couple of days ago, or
(c) refraining from reaching for the cup was a spontaneous decision.
According to the explicit definition held by neuroscientists, my
decision to veto is voluntary in all these cases: the decision is not the
direct result of an external cue, it is generated internally, and I experi-
ence control over my decision — I feel that I could have reached for it
anyhow had I decided to do so.
However, in the experiments the subjects had to decide to veto their
initial urge or intention or not in circumstances that were much more
restricted than these descriptions allow. The subjects had to wait for
an intention or urge to move to emerge and then decide spontaneously,
without pre-planning, to veto or not. No external factors were present
or changed to make sure that they could not causally contribute to the
decision that was made.6 Furthermore, and this is where the
discrepancy between the explicit definition and the implicit operation-
alization becomes clear, the experiments are conducted in such a way
that no internal cues or factors were manipulated, measured, or
allowed to play a role either. There was no goal to be reached or

acting in socially desirable ways (Filevich, Kühn and Haggard, 2013), or resisting
criminal activities (Parkinson and Haggard, 2014).
6 Kühn, Haggard and Brass (2009) did integrate external cues in their experiment,
because previous studies on the veto ‘lacked the context of reasons for performing and
inhibiting action that generally exist outside the laboratory’ (ibid., p. 2835). However,
the subjects were not allowed to decide on the basis of these reasons whether to veto or
not. Later in the paper it will become clear why this is important.
 14 L. ASMA

desire to be met by making the movement or deciding to veto, and the

subjects were not allowed to let goals or desires play a role. They were
told not to pre-plan what they were going to do, and to decide
spontaneously. Furthermore, the frequency of vetoing and not vetoing
was supposed to be more or less equal, and subjects were not allowed
to use strategies to achieve this equality. Thus, the researchers
attempted to exclude all possible influences on the decision from the
experimental setting, external and internal, in the sense that goals,
desires, or beliefs were not allowed to contribute to the decision that
was made. That means that only (c) counts as a voluntary veto
decision.
Schüür and Haggard (2011) refer to these actions and decisions as
For personal use only -- not for reproduction

they are actually studied in the experiments as underdetermined,

which I also use here for these kinds of decisions.7 As they (ibid., p.
Copyright (c) Imprint Academic 2017

1699) state, this operationalization seems to be based on our experi-

ence that sometimes ‘I’ am the cause of my decisions and actions,
instead of internal or external factors. By leaving all cues or factors
that could count as reasons for action out of the experimental setting,
the decision emerges from the self, ‘a source other than external or
internal cues’ (ibid., p. 1699). In giving subjects such a task, it seems
that the researchers tried to capture the intuition that sometimes I can
decide what to do without being influenced. Thus only spontaneous
decisions that are not based on any reason, either internal or external,
count as voluntary on this definition. And this is the kind of decision
that is the object of study in Libet-style experiments.
To conclude, the descriptions and examples in the literature in
principle allow internal factors such as beliefs or desires to play a role
in the decision the subjects have to make. However, the explicit
definition and implicit operationalization differ. This means that
Schlosser (2014) is not capturing the implicit operationalization that is
in fact used in the experiments. The neuroscientists in the field do not
distinguish between internal and external reasons for actions in their
experiments, but the experiments are set up in such a way that the
subjects are not allowed to take external or internal factors into
account.

7 Schüür and Haggard (2011) also criticize this description. I use the remarks from their
paper only to clarify what underdetermined decisions are, and what the rationale is
behind operationalizing voluntariness this way.
 THERE IS NO FREE WON’T 15

Free Won’t: Conscious Causation or Initiation?

In this section I examine the definition of freedom that is central in
conclusions drawn from Libet-style experiments. Based on the conclu-
sions drawn and the claims made in the literature, I show that neuro-
scientists in the field do not think that a causal role for conscious
processes is sufficient for a decision to be free. Instead, the decision
has to be initiated consciously, and should not be part of a causal
chain that started unconsciously.
The results of Libet-style experiments are sometimes used as
evidence for the claim that conscious processes play no causal role at
all. For example, Libet concludes that the brain has already made the
decision when to move (Libet, 1985, p. 536), and that our conscious
For personal use only -- not for reproduction

intention to move merely ‘bubbles up’ from our brain (Libet, 1999, p.
Copyright (c) Imprint Academic 2017

54). This finding that the conscious intention is ‘a consequence or

correlate of neural preparation of action’ is interpreted as a threat to
free will (e.g. Haggard and Libet, 2001, p. 46). It is concluded from
this that our conscious intention has no causal role to play; the neural
process is enough for the movement to occur. Wegner (2002, p. 58)
argued, partly based on these studies, that we might experience that
we consciously initiate movements, but in fact unconscious brain
activity causes both our actions and our conscious intentions. This
way, conscious intentions are conceived as inefficacious end products
of brain activity that do not play a role in bringing about our actions.
Because of this, it is important to argue that this conclusion, that the
experiments show that conscious intentions are never causally
effective, does not follow from the experiments as for example Mele
(2009) argues.
However, from the recent discussions and experiments on free
won’t, it is clear that even if conscious processes play a role, this is
not taken to be sufficient to conclude that the voluntary decision was a
free decision as well. If we assume that subjects do as they are told in
the experiments, conscious processes play a role in the voluntary veto
decision. Subjects first have to become conscious of the urge or
intention to move before they can decide to veto, they have to have
‘the experience of being about to commit an action’ (Filevich, Kühn
and Haggard, 2012, p. 1107). Furthermore, Filevich et al. explicitly
state that there is a conscious decision involved in the experiment:
‘[t]he current state of the brain appears to influence the conscious
decision to act or inhibit/delay, rather than vice versa’ (Filevich, Kühn
and Haggard, 2013, p. 8). Thus, if it were sufficient for a decision to
 16 L. ASMA

be free if conscious processes played a causal role, the voluntary veto

decision would be free. However, it turns out that it is not sufficient.
Even if conscious processes played a causal role, ‘a stronger rejection
of “free won’t” could come from actually showing that a decision to
act or not can be driven by a preceding, presumably unconscious
neural activity’ (ibid., p. 10). If there is this preceding unconscious
neural activity, the brain would still ‘decide’ whether we are going to
act on our initial conscious intention or not. Based on the results of
Filevich, Kühn and Haggard (2013), the veto decision is understood as
initiated by unconscious neural activity and not free, even though con-
scious awareness of the intention or urge to move has to play some
role in the subsequent decision that is made. Neuroscientists in the
For personal use only -- not for reproduction

field seem to assume that for a decision to be free it has to be an

uncaused cause, a decision made by a conscious self that exists
Copyright (c) Imprint Academic 2017

separate from the brain, for which no predictive neural correlates may
be found. Filevich et al. capture this intuition by stating that their
‘results suggest that an important aspect of “free” decisions to inhibit
can be explained without recourse to an endogenous, “uncaused” pro-
cess: the cause of our “free decisions” may at least in part, be simply
the background stochastic fluctuations of cortical excitability’ (ibid.,
p. 10).
Thus, the veto decision would only be free, according to neuro-
scientists’ definition of freedom, if the decision is consciously
initiated, which means that it cannot be predicted on the basis of pre-
ceding unconscious neural processes. That means that arguing that
conscious processes play a causal role, as Mele (2009) does, does not
fully target the stronger assumption that underlies the conclusion that
is drawn from the experiment.

What Underdetermined, Consciously

Initiated Decision Making Amounts To
In this section my aim is not primarily to criticize the definitions of
voluntariness and freedom that I put forth in the previous sections.
Rather, my aim is to argue that, because of these underlying
assumptions and this experimental setting, the conclusion that the
brain ‘decides’ is almost unavoidable. I start this section by showing
that those who take the findings from Libet-style experiments
seriously, philosophers, and ‘the folk’ think that (consciously) acting
for and responding to reasons is central for free will. After that I
provide three arguments for why this threat, that actions are not the
 THERE IS NO FREE WON’T 17

result of reasons or reasoning processes and that instead the brain

‘decides’, emerges from the experiments precisely because it is
impossible for the subjects to be conceived as acting for reasons. As a
result, only spontaneous decisions that are not based on any reason nor
caused in any way would count as free.
In general, Libet-style experiments, also the one conducted by
Filevich, Kühn and Haggard (2013) on free won’t, are taken to be
threatening to free will because they seem to show that our conscious
agency is ‘bypassed’. The experiments seem to show that we do not
consciously decide what to do, but in fact our brain made the decision
for us. For example Caruso (2012) takes the finding that decisions and
actions have unconscious neural precursors ‘to be a serious threat to
For personal use only -- not for reproduction

our intuitive sense of free will’ (p. 194). He argues that our intuitive,
common-sense conception of free will involves a central role for con-
Copyright (c) Imprint Academic 2017

sciousness, mainly because consciousness is generally associated with

our abilities as agents to deliberate and guide our actions and
decisions. He states that ‘[f]or there to be free will, there must be a
difference between what happens to an individual and an individual’s
active participation in the process of deliberation, choice, and action
initiation’ (ibid., p. 205). Since Libet-style experiments are taken to
show that our conscious decisions are initiated unconsciously, it seems
that these abilities cannot and do not make a difference to the decision
that is made. Also free will theorists take the ability to respond to and
act for reasons to be a central aspect of free will (see for example
Mele, 2003, p. 212, and Schlosser, 2014). Furthermore, this also
seems to be what ‘the folk’ worry most about. For example Nahmias
(2006, p. 217) states on the basis of his surveys on folk intuitions
about free will that ‘any thesis that suggests our actions are caused by
processes that bypass our conscious mental life — that bypass us, as it
were — is intuitively threatening to freedom and responsibility’. Thus,
the fact that our ability to act for and respond to reasons that we
generally associate with consciousness may not play a role in our
actions because unconscious neural activity in fact caused it, is what
we take to be threatening about these experiments.
However, concluding that these experiments show that our con-
scious agency is bypassed in situations other than these experiments is
problematic for a couple of reasons. Remember that, because of the
two central assumptions underlying the experiments, subjects had to
decide to veto or not in a situation in which (a) the only factor
involved in the decision was the urge to make a preset movement
based on the researchers’ instructions (and perhaps an urge or
 18 L. ASMA

inclination to veto or not) and (b) the decision to veto or not should be
consciously initiated, or, in other words, not be predictable on the
basis of unconscious brain activity. It is important to realize that this
setting almost unavoidably leads to the conclusion that the brain
‘decides’ and that conscious deliberation is bypassed. I will provide
three arguments for this claim.
First, because the subjects had to make voluntary veto decisions of
the underdetermined kind, there was nothing to consciously deliberate
about in the experimental setting. The experiment was set up in such a
way that subjects did prepare to move and probably did acquire a
conscious urge or intention to move.8 Since the goal was to exclude all
external and internal cues from the experimental setting, that urge to
For personal use only -- not for reproduction

move is the only factor that may play a role in the decision to veto or
not. This means that there is simply nothing to deliberate about and
Copyright (c) Imprint Academic 2017

this ability could not make a difference to the outcome.9

Second, if subjects did use these abilities, even though they were not
supposed to because the decision should be underdetermined, the
related brain activity would not be interpreted as reflecting conscious
decision making or deliberation, but as unconscious neural precursors
of the decision. If I was a subject in these experiments and I found it
hard to decide whether to veto or not, I may invoke a certain rhythm
in my responses, for example on the basis of musical notes of my
favourite song (‘C’ is veto, ‘D’ is not, etc.). Obviously, related brain
activity can be found and it might be possible to predict my decisions
on the basis of that. If that were the case, this brain activity would be
interpreted as an unconscious precursor of my decision, while in fact it
was the result of my conscious efforts to remember the musical notes
of the song and make decisions on the basis of that. But since the sub-
jects are asked not to do this and they do not report on these processes,
the experimenters assume that this neural activity is unrelated to con-
scious processes. Thus, if the subjects reasoned about what to do and
exercise abilities that we associate with free will, this would be inter-
preted as evidence that the decisions are not free. Even more, on the

8 As said, Mele (2008) argued that this might not be what actually happens, see footnote
3.
9 This point, often put as Libet-style experiments being about the liberty of indifference,
has been made before in the literature, see for example Mele (2009, p. 84). However,
my aim is not mainly to criticize this definition of freedom directly, but to show that
thinkers who take the findings of Libet-style experiments seriously demand something
that will not occur in this experimental set-up.
 THERE IS NO FREE WON’T 19

basis of the definition of voluntariness that is held in the experiments,

even if it was acknowledged that the subject decided on the basis of
the song, the decision would not count as free. It would be a decision
that is based on internal factors and therefore not voluntary nor free.
Third, by taking away all factors that the decision to veto or not
could be based on it is not unlikely that subjects more or less let their
brain make the decision. By not allowing any possible influence on
the decision that is made, no factor is left in the experimental setting
other than brain fluctuations and possibly related urges or inclinations.
This is also what the researchers themselves state: ‘[i]n the absence of
any external cause to act rapidly or inhibit, we hypothesized that some
other factors, such as transient fluctuations in subjects’ brain states,
For personal use only -- not for reproduction

may be relevant to their decision in this case’ (Filevich, Kühn and

Haggard, 2013, pp. 1–2). Conscious processes are interpreted as
Copyright (c) Imprint Academic 2017

playing a causal role only if they change something in the outcome of

the experiment. But since the subjects were not allowed to take any
factors into account in this decision, it is difficult to conceive of a way
in which subjects could ‘disrupt’ the relations between the predictive
brain activity and the decision if there is only this urge to decide and
act on.
Because of these limitations, the question is what the result of the
experiment should be in order to conclude that the veto decision is
free. This would not have much to do with conscious deliberation, but
would amount to a spontaneous decision that is not based on reasons
nor caused in any way. The conclusion that the conscious self decides
instead of the brain would be drawn only if no previous brain activity
was found that correlated with the conscious decision that was made
and no reasons were taken into account. Then, conscious intentions
have to appear out of nowhere, and even then it is very likely that
correlations between the intention and some preceding brain activity
can be found.
To conclude, the experimental setting of Libet-style experiments
forces subjects to decide to veto or not without input to actually
deliberate about what to do. Furthermore, if they did deliberate about
what to do, it would not be interpreted as such by the researchers.
Because of this, conducting experiments with these definitions of
voluntariness and freedom in place almost unavoidably leads to the
conclusion that the brain ‘decided’ and not the conscious self, and it
leads to a distorted picture of the difference that our abilities as con-
scious agents can actually make. Alternative definitions of voluntari-
ness and freedom are necessary for neuroscientific experiments to
 20 L. ASMA

provide a valuable contribution to the debate on free will and free

won’t.

Alternative Definitions of Voluntariness

and Freedom in the Neuroscientific Literature
The question is whether voluntary and/or free decisions can be defined
in such a way that they can be studied experimentally in a way that the
experiments contribute to our understanding of free will. This question
is not new, since also neuroscientists in the field think defining
voluntariness as underdetermined is problematic (e.g. Nachev and
Husain, 2010; Schüür and Haggard, 2011). In the literature on this
topic, (at least) two alternative conceptions of voluntariness have been
For personal use only -- not for reproduction

proposed: operant action and acting on the basis of complex integra-

Copyright (c) Imprint Academic 2017

tion. The central idea in both definitions is that voluntary actions and
decisions should not be caused directly by external cues or at least
‘not uniquely specified by external cues’ (Schüür and Haggard, 2011,
p. 1698). In this section I examine these proposals and point out some
limitations. Furthermore, I elaborate on the fact that the definition of
freedom seems to be generally accepted.
In the case of operant action, distinguished from respondent action,
the criterion is that the subject is not directly responding to external
cues. Passingham and Lau (2006) describe free choice as ‘choices that
are not imposed on the subject by an external agent’ (p. 53). In a
similar vein, Filevich, Kühn and Haggard (2012, p. 1108) describe
operant veto decisions as decisions that are internally generated: ‘[b]y
definition, the process or signal that cancels or inhibits the action is
not the result of any external signal or instruction, but is generated
internally by the subject themselves.’ There are different ways to
interpret these descriptions, but the idea is that when internal factors
bring about the actions, they depend ‘on wide integration of informa-
tion beyond the current stimulus, such as memory of previous experi-
ence and predictions of future outcomes’ (ibid., p. 1116). A different
proposal comes from Schüür and Haggard (2011). They state that
voluntary actions (and decisions) should be seen as ‘the motor con-
sequences of processing and integrating large numbers of qualitatively
different types of input’ (ibid., p. 1702). For example, a soccer player
has to integrate many inputs and strategically respond to them during
a game. They propose that in order to study voluntary action and
decision making in a better way, what is needed is not a reduction of
the number of external inputs, but rather an increase of the number of
 THERE IS NO FREE WON’T 21

different types of inputs (ibid., pp. 1702–3). Then, ‘everything’ is the

cause of self-generated actions and decisions.
However, these definitions still allow for many different interpreta-
tions. Does that mean that all the actions and decisions that are sponta-
neously generated from my brain are free? Or is it limited to actions
and decisions that have causes internal to the conscious self? In the
first case, also behaviours or inhibitions of actions that are the result
of seizures are included. In the last case many questions remain: does
that mean that we have to be conscious of the factors that we base the
choice on, or is it necessary that conscious deliberation took place?
For example, I can be conscious of my heartbeat, but that does not
mean that I can make an (effective) choice about it. Furthermore, my
For personal use only -- not for reproduction

conscious deliberation can be influenced by factors that I am not

aware of. How should we treat these cases? And why can’t we allow
Copyright (c) Imprint Academic 2017

for ‘simple’ decisions, which are not the result of complex integration,
to be free?
It seems that the general problem with these definitions is that the
researchers do not make a distinction between actions and decisions
that are the result of reasons, and movements that are merely caused.
Philosophers generally consider voluntary actions and decisions to be
special because we, as agents, play an active role in the outcome and
what we decide and do is not merely a product of causal ‘blind’
forces. We are able to take reasons into account, deliberate about
them, and because of that we can effectively respond to facts and
circumstances in a way that we find reasonable or justified. We want
to make a difference between hurting someone as a result of an acci-
dent, and hurting someone because of a reason, e.g. disliking the
person. The central point is that, in cases in which we act for reasons,
our actions can be rationalized from our own point of view (Schlosser,
2014, p. 250). However, the alternative definitions of what it means to
act voluntarily that neuroscientists in the field provide do not capture
this essential distinction between ‘blind’ causation and acting for
reasons. In order to do this, the relative contribution of internal and
external factors does not seem to explain this difference, as also
Schlosser (2014) argued in his paper, and it is unclear how enhancing
the complexity of the decision contributes to this. Thus, the proposal
would be to study actions that are done for reasons.
Furthermore, it is interesting that the understanding of free will is
not questioned in the neuroscientific literature in the field. In the
current conception in which conscious initiation is central, it is
assumed that for a decision to be free it has to be initiated by a
 22 L. ASMA

conscious self that exists separate from the brain. However, they do
not argue for their position nor give an account of how this would
work, and/or how this would make free will and conscious action
control possible. Furthermore, their understanding of free will does
not seem to track folk intuitions about free will (see, for an overview
of conducted surveys, Mele, 2012). The burden of proof seems to be
on those that defend a supernatural or substance dualist account of
free will, which are in this case the neuroscientists that study it.
Thus, alternative definitions of voluntariness and freedom are
necessary for neuroscientific experiments to provide a valuable contri-
bution to the debate on free will and free won’t. In order to do so, they
should at least acknowledge that a central aspect of free will is
For personal use only -- not for reproduction

whether the agent acted for reasons. Moreover, neuroscientists in the

field seem to assume that conscious free will can only exist if some
Copyright (c) Imprint Academic 2017

kind of dualism is true, while a role of consciousness in freedom could

also be accommodated on the basis of scientifically accepted theories
of consciousness. For now, the question to what extent neuroscience
can contribute to our understanding of free will is open. At the very
least, conclusions drawn about the role of conscious processes or our
agency based on these experiments need to be put off, because the
conclusions are drawn based on experiments in which problematic
assumptions play a large role.
Acknowledgments
This publication was made possible through the support of a grant
from the Templeton World Charity Foundation. The opinions
expressed in this publication are those of the author and do not
necessarily reflect the views of the Templeton World Charity Founda-
tion. An earlier version of this paper was given at the SINe Confer-
ence, GCTP14, CCN Conference on Free Will, and OZSW Confer-
ence 2014, and I would like to thank the members of the audience for
their questions and remarks on the paper. Also, I would like to thank
Leon de Bruin, Hans van Eyghen, Marcelo Fischborn, Gerrit Glas,
Naomi Kloosterboer, Kelvin McQueen, Al Mele, Rik Peels, Jeroen de
Ridder, Scott Robbins, Emanuel Rutten, Jojanneke Vanderveen,
David Widerker, Jan Willem Wieland, and René van Woudenberg for
their helpful comments on earlier drafts of this paper.

References
Brass, M. & Haggard, P. (2007) To do or not to do: The neural signature of self-
control, Journal of Neuroscience, 27 (34), pp. 9141–9145.
 THERE IS NO FREE WON’T 23

Caruso, G.D. (2012) Free Will and Consciousness: A Determinist Account of the
Illusion of Free Will, Lanham, MD: Lexington Books.
Filevich, E., Kühn, S. & Haggard, P. (2012) Intention inhibition in human action:
The power of ‘no’, Neuroscience and Biobehavioral Reviews, 36, pp. 1107–
1118.
Filevich, E., Kühn, S. & Haggard, P. (2013) There is no free won’t: Antecendent
brain activity predicts choices to inhibit, PLoS ONE, 8 (2), e53053.
Fischborn, M. (2016) Libet-style experiments, neuroscience, and libertarian free
will, Philosophical Psychology, DOI: 10.1080/09515089.2016.1141399.
Haggard, P. (2008) Human volition: Towards a neuroscience of will, Nature
Reviews Neuroscience, 9, pp. 934–946.
Haggard, P. & Eimer, M. (1999) On the relation between brain potentials and the
awareness of self-generated movements, Experimental Brain Research, 126, pp.
128–133.
Haggard, P. & Libet, B. (2001) Conscious intention and brain activity, Journal of
For personal use only -- not for reproduction

Consciousness Studies, 8 (11), pp. 47–63.

Kühn, S., Haggard, P. & Brass, M. (2009) Intentional inhibition: How the veto-
Copyright (c) Imprint Academic 2017

area exerts control, Human Brain Mapping, 30 (9), pp. 2834–2843.

Libet, B. (1985) Unconscious cerebral initiative and the role of conscious will in
voluntary action, The Behavioral and Brain Sciences, 8, pp. 529–566.
Libet, B. (1999) Do we have free will?, Journal of Consciousness Studies, 6 (8–9),
pp. 47–57.
Mele, A.R. (2003) Motivation and Agency, New York: Oxford University Press.
Mele, A.R. (2008) Proximal intentions, intention-reports, and vetoing, Philo-
sophical Psychology, 21 (1), pp. 1–14.
Mele, A.R. (2009) Effective Intentions: The Power of Conscious Will, New York:
Oxford University Press.
Mele, A.R. (2012) Another scientific threat to free will?, The Monist, 95 (3), pp.
422–440.
Nachev, P. & Husain, M. (2010) Action and the fallacy of the ‘internal’: Comment
on Passingham et al., Trends in Cognitive Sciences, 14 (5), pp. 192–193.
Nahmias, E.A. (2006) Folk fears about freedom and responsibility: Determinism
vs. Reductionism, Journal of Cognition and Culture, 6 (1–2), pp. 215–237.
Parkinson, J. & Haggard, P. (2014) Subliminal priming of intentional inhibition,
Cognition, 130, pp. 255–265.
Passingham, R.E. & Lau, H.C. (2006) Free choice and the human brain, in Pockett,
S, Banks, W.P. & Gallagher, S. (eds.) Does Consciousness Cause Behavior?,
Cambridge, MA: MIT Press.
Schlosser, M.E. (2014) The neuroscientific study of free will: A diagnosis of the
controversy, Synthese, 191, pp. 245–262.
Schüür, F. & Haggard, P. (2011) What are self-generated actions?, Consciousness
and Cognition, 20, pp. 1697–1704.
Soon, C.S., Brass, M., Heinze, H.-J. & Haynes, J.-D. (2008) Unconscious
determinants of free choice in the human brain, Nature Neuroscience, 11 (5), pp.
543–545.
Wegner, D.M. (2002) The Illusion of Conscious Will, Cambridge, MA: MIT Press.

Paper received June 2016; revised October 2016.